Reporting of Sociodemographic Variables in Randomized Clinical Trials, 2014-2020

Disease burden, health outcomes, and the effectiveness of interventions are associated with sociodemographic variables, such as race and ethnicity, income, and educational level. Understanding these social determinants of health and evaluating the effectiveness of health interventions among different sociodemographic groups is an essential approach to addressing health inequities.¹

Randomized clinical trials (RCTs) provide the best available method to examine the causal effects of interventions in health care research. Such trials can serve the needs of racialized, low-income, and other marginalized and underserved communities only if sociodemographic variables and social determinants of health are measured and reported. It is unclear how frequently and how thoroughly the sociodemographic characteristics of participants are reported in RCTs. The objective of this study was to identify the frequency of reporting of sociodemographic variables in RCTs published in 5 high-impact health journals.

A cross-sectional study was conducted of 10% of all RCTs published in 5 journals—New England Journal of Medicine, JAMA, The BMJ, The Lancet, and Annals of Internal Medicine—from January 1, 2014, through July 31, 2020. Every 10th article was retained from the search, starting with the most recently published.² Each article was assessed by one of 4 reviewers (including G.N.) to identify the study’s medical discipline, intervention categories, study objective, and funding source and the sociodemographic variables used to characterize the population. Data queries were resolved by consensus. This study fell outside the scope of ethics committee review according to Canada’s Tri-Council Policy Statement on the ethical conduct for research involving humans. We followed applicable elements of the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. The findings were tabulated to describe the sociodemographic variables reported in the included RCTs. All data analysis was conducted with Google Sheets spreadsheet software (Google).

During the study time frame, 2351 RCTs were published in the selected journals; 237 articles were randomly selected for review (Table 1). The most commonly reported sociodemographic variables were age (234 of 237 articles [98.7%]), any descriptor of sex or gender (234 of 237 [98.7%]), and race/ethnicity (115 of 237 [48.5%]) (Table 2). All other sociodemographic variables were reported in less than 15% of trials. Educational level or literacy was the next most commonly reported variable (34 articles [14.3%]). Income or socioeconomic status were rarely reported (14 [5.9%]). Three trials (1.3%) concerned interventions targeting social determinants of health, and 11 studies (4.6%) included participants based on a social determinant of health. Twenty-five studies (10.5%) were conducted in an all-male or all-female population; all of these trials studied diseases that affect only 1 sex (eg, prostate cancer, induction of labor). Six articles (2.5%) reported gender. No studies reported nonbinary gender descriptors.

The findings of this cross-sectional study show that sociodemographic characteristics of study populations are minimally reported in RCTs published in high-impact journals. Descriptors of participant educational level or literacy were reported in only 14.3% of trials, and descriptors ofhousing, language, access to health care, or employment status appeared in less than 5% of studies.

White, male, and wealthy individuals are more often recruited and retained in RCTs compared with the general population or the population affected by the condition under investigation.^3-5 The failure to gather or report on sociodemographic and social determinants of health obscures inequities in trial enrollment and outcomes. This threatens the applicability of trials, raises ethical concerns, and inhibits hypothesis building, subgroup analyses, and evidence syntheses.

This study has limitations. The articles included in the study represent a 10% sample of high-impact English-language publications and may not be representative of all RCTs. In addition, trialists may gather sociodemographic data but not report them.

All trial stakeholders share the responsibility to ensure that RCTs serve the health needs of the communities they are intended to serve. Randomized clinical trials cannot contribute to correcting health inequities if sociodemographic and social determinants are not measured and not reported. The Consolidated Standards of Reporting Trials–Equity 2017 statement calls for reporting sociodemographic variables but is only applicable to trials concerning health equity.⁶ For all trials, regulatory standards could help ensure that sociodemographic data are gathered and reported, whereas reporting guidelines and editorial standards could ensure that sociodemographic data are reported transparently.

Accepted for Publication: March 20, 2021.

Published: June 2, 2021. doi:10.1001/jamanetworkopen.2021.10700

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Orkin AM et al. JAMA Network Open.

Corresponding Author: Aaron M. Orkin, MD, MSc, MPH, Li Ka Shing Knowledge Institute, Unity Health Toronto, 250 Yonge St, Toronto, ON M5B 1W8, Canada (aaron.orkin@mail.utoronto.ca).

Author Contributions: Dr Orkin and Ms Nicoll had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Orkin, Persaud, Pinto.

Acquisition, analysis, or interpretation of data: Orkin, Nicoll, Persaud.

Drafting of the manuscript: Orkin.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Nicoll.

Administrative, technical, or material support: Orkin, Nicoll, Pinto.

Supervision: Orkin, Nicoll, Pinto.

Conflict of Interest Disclosures: Dr Orkin reported being supported by the Canadian Institutes of Health Research Postdoctoral Fellowship Program and the Department of Family and Community Medicine, Faculty of Medicine, University of Toronto. Dr Persaud reported receiving grants from the Canada Research Chairs Program and the Canadian Institutes of Health Research outside the submitted work. Dr Pinto reported being supported as a clinician-scientist by the Department of Family and Community Medicine, Faculty of Medicine, University of Toronto; the Department of Family and Community Medicine, St Michael’s Hospital; the Li Ka Shing Knowledge Institute, and St Michael’s Hospital; by a fellowship from the Physicians’ Services Incorporated Foundation; and as the associate director for clinical research at the University of Toronto Practice-Based Research Network. Dr Pinto reported receiving honoraria for presentations at Queen’s University (2010), the University of Saskatchewan (2012), Mount Sinai Hospital (2012), the Toronto Reference Library (2016), the Law Society of Ontario (2016), the Japan Network of Health Promoting Hospitals and Health Services (2018), the Joint Centre for Bioethics, University of Toronto (2019), McMaster University (2020), and Ghent University, Belgium (2020). Dr Pinto also reported serving as an unpaid scientific advisor to a start-up company, Mutuo Health Solutions. No other disclosures were reported.

Additional Contributions: We thank Ross Upshur, MD, MA, MSc, for his contributions to developing this study and Jordynn Klein, BAS, Jack Iglar, and Muhamad Ally, BSc (all of the University of Toronto) for their assistance with reviewing citations and data abstraction. None of these individuals were compensated for their contributions.